The most successful of all carbohydrate pharmaceuticals so far have been the carbohydrate-based, antibacterial vaccines [1]. The basis of using carbohydrates as vaccine components is that the capsular polysaccharides and the O-specific polysaccharides on the surface of pathogenic bacteria are both protective antigens and essential virulence factors. The first saccharide-based vaccines contained capsular polysaccharides of Pneumococci: in the United States a 14-valent vaccine was licensed in 1978 followed by a 23-valent vaccine in 1983. Other capsular polysaccharides licensed for human use include a tetravalent meningococcal vaccine and the Vi polysaccharide of Salmonella typhi for typhoid fever. The inability of most polysaccharides to elicit protective levels of anti-carbohydrate antibodies in infants and adults with weakened immune systems could be overcome by their covalent attachment to proteins that conferred T-cell dependent properties [2]. This principle led to the construction of vaccines against Haemophilus influenzae b (Hib) [3] and in countries where these vaccines are routinely used, meningitis and other diseases caused by Hib have been virtually eliminated [4]. Extension of the conjugate technology to the O-specific polysaccharides of Gram-negative bacteria has provided a new generation of glycoconjugate vaccines that are undergoing various phases of clinical trials [5].
Escherichia coli O157:H7, an emerging infectious agent, was first recognized as a human pathogen in 1983 [6]. Diseases caused by this pathogen have subsequently been recognized worldwide [7]. Infection with E. coli O517 causes a spectrum of illnesses with high morbidity and mortality, ranging from watery diarrhea to hemorrhagic colitis and the extraintestinal complication of hemolytic-uremic syndrome (HUS). HUS can lead to acute renal failure requiring dialysis, and in children and infants this complication has a considerable mortality. In some studies, E. coli O517 was the most common cause of dysentery in patients seen in hospital clinics [8].
E. coli strains associated with HUS produce at least one toxin identical to the exotoxin of Shigella dysenteriae serotype 1, referred to herein as Shiga toxin 1 (Stx1). This toxin has been variously referred to in the literature as Vero cytotoxin 1 (VT1), Shiga-like toxin 1 (SLT-I), and Shiga toxin 1(Stx-I or Stx1). In some cases a second toxin (variously referred to as VT2, SLT-II, Stx-II, or Stx2), structurally and functionally related to Stx1 and having a cross-reactive A subunit, is also produced. Infection with Stx-producing organisms has been correlated with HUS, and E. coli O517:H7 is a common serotype that produces these toxins. However, strains of E. coli O517 without Stx have been isolated from patients with hemorrhagic colitis.
The pathogenicity of E. coli O517 has been compared to that of Shigella dysenteriae type 1 [9, 10]. Both E. coli O517 and S. dysenteriae type 1 secrete almost identical exotoxins (Stx1 or Stx2) and cause bloody diarrhea, with its complications, only in humans. Antibiotic treatment does not ameliorate the course of enteritis caused by E. coli O517, and it may in fact increase the incidence of HUS caused by E. coli and S. dysenteriae type 1 [11,12]. Unlike S. dysenteriae type 1, which is confined to humans, E. coli O157:H7 lives in cattle and in other domesticated animals without causing symptoms. The feces of infected animals serve as a source of E. coli O517 infection in humans, through contamination of drinking water and meat.
Most adults have low or nondetectable levels of serum antibodies to E. coli O517 O-SP and to Shiga toxins. High levels of O-SP antibodies and low or nondetectable levels of antitoxin are regularly found following infection with E. coli O517 and the subsequent complication HUS. It is not known whether immunity follows infection with this pathogen.
Although there is no consensus on the host factors that might confer immunity to E. coil O157, the O-specific polysaccharide portion of the lipopolysaccharides of the similar genus Shigella have emerged as possible protective antigens [13,14]. These polysaccharides were shown to be essential for the virulence of Shigella, and it is now well-established that the protection is serotype specific. Since each serotype is characterized by a distinct O-specific polysaccharide, it is fair to say that protection against E. coli O517 is also O-specific polysaccharide specific. The safety and immunogenicity of a protein conjugate of the O-specific polysaccharides of S. sonnei, S. flexneri 2a, and S. dysenteriae type 1 has been demonstrated in human volunteers, and preliminary clinical trials have established the efficacy of these vaccines [9, 15, 16, 17].
The immunogenicity of saccharides, alone or as protein conjugates, is related to several variables: 1) species and the age of the recipient; 2) molecular weight of the saccharide; 3) density of the saccharide on the protein; 4) configuration of the conjugate (single vs. multiple point attachment); and 5) the immunologic properties of the protein.
Because high molecular weight polysaccharides can induce the synthesis of antibodies from B-cells alone, they are described as T-independent antigens. Three properties of polysaccharides are associated with T-independence; 1) their repetitive polymeric nature, which results in one molecule having multiple identical epitopes; 2) a minimum molecular weight that is related to their ability to adhere to and cross-link membrane-bound IgM receptors, resulting in signal transduction and antibody synthesis; and 3) resistance to degradation by mammalian enzymes. Most capsular polysaccharides are of comparatively high molecular weight (≧150 kD), and elicit antibodies in older children and in adults but not in infants and young children. O-SPs are of lower molecular weight (≧100 kD), and may be considered to be haptens because they combine with antibody (are antigenic) but do not elicit antibody synthesis (are not immunogenic). The immunogenicity of O-SPs as conjugates may be explained by two factors: 1) the increase in molecular weight that allows the O-SP to adhere to a greater number of membrane-bound IgM and induce signal transduction to the B-cell; and 2) their protein component, which is catabolized by the O-SP stimulated B cell resulting in a peptide-histocompatibility II antigen signal to T cells.
Synthesis of conjugates for use as vaccines in humans has special considerations. LPS is not suitable for parenteral administration to humans because of toxicity mediated by the lipid A domain. Usually, O-SP is prepared by treatment of LPS with either acid or hydrazine in order to remove fatty acids from lipid A. The resultant products retain the core region and the O-SP with its heterogeneous range of molecular weights (Mr). Conjugates are prepared by schemes that bind the carrier to the O-SP at multiple sites along the O-SP, or attempt to activate one residue of the core region.
In the case of E. coli O517, vaccine development has been hindered because there is little information about mechanisms of immunity [9], and there are no valid animal models for diseases caused by E. coli O517[10].
There have been some efforts to date to attempt to obtain effective vaccine compositions against E. coli. See, e.g., Cryz et al. (U.S. Pat. No. 5,370,872), which describes the isolation of O-SP derived from LPS of 12 serotypes of E. coli and their covalent linkage to P. aeruginosa toxin A as a carrier protein [18]. The twelve monovalent conjugates were combined to form a polyvalent vaccine, which was described as being safe and immunogenic in both rabbits and humans when administered by injection. An antibody response to both the O-SP and toxin A moieties was reported, and protection of rabbits against E. coli sepsis was demonstrated upon passive immunization with the resulting IgG antibodies. However, neither bactericidal activity of the antibodies nor protection after vaccination with the conjugates was shown, and antibodies against E. coli strain O517 and protection against E. coli O517 infection are not mentioned.
Because anti-LPS or anti-O-SP antibody-mediated protection is likely to be serotype-specific, it is unlikely that the polyvalent vaccine described in U.S. Pat. No. 5,370,872 would induce a significant level of antibodies against E. coli O517 O-SP or LPS. There remains a need, therefore, for compositions and methods of inducing a significant level of antibodies against E. coli O517. There also remains a need compositions and methods for inducing antibodies which have bactericidal activity against E. coli O517, and which also prevent or ameliorate HUS.